Seismic reflection images of active faults on New Zealand's South Island

A.G. Green1, F.M. Campbell1, A.E. Kaiser1, C. Dorn1, S. Carpentier1, J.A. Doetsch1, H. Horstmeyer1, D. Nobes2, J. Campbell2, M. Finnemore2, R. Jongens3, F. Ghisetti4, A.R. Gorman4, R.M. Langridge5, A.F. McClymont6 1. Institute of Geophysics, ETH, Zrich, CH-8092, Switzerland 2. Department of Geological Sciences, University of Canterbury, Christchurch, New Zealand 3. Dunedin Research Centre, GNS Science, Private Bag 1930, Dunedin, 9054, New Zealand 4. Department of Geology, University of Otago, Dunedin, 9054, New Zealand 5.GNS Science, P.O. Box 30-368, Lower Hutt, New Zealand 6. Department of Geoscience, University of Calgary, Calgary, T2N1N4, Canada Abstract New Zealand is located along the boundary between the Australian and Pacific plates. Although there are numerous faults associated with this plate boundary setting, few have ruptured during the nearly 200 years of European settlement. Yet, paleoseismology provides clear evidence of relatively recent activity on many of them. Knowledge of the shallow structure and other characteristics of these faults is important for understanding the related seismic hazard and risk. Key properties of faults that produce infrequent large earthquakes are usually determined or inferred from paleoseismological investigations of surface outcrops, geomorphology, trenches, and boreholes. In an attempt to improve our knowledge and understanding of active faults beyond the reach of conventional paleoseismological methods (i.e., deeper than a few meters), we have acquired high-resolution seismic reflection and ground-penetrating radar (GPR) data across the following three fault systems on New Zealand's South Island: (i) a northern section of the transpressive Alpine Fault zone, (ii) numerous reverse faults hidden beneath the very young sediments that cover the northwest Canterbury Plains, and (iii) a critical portion of the reverse Ostler Fault zone in the south-central part of the Island. After subjecting our data to diverse processing procedures, the resultant seismic and GPR sections provide vivid images of the target structures. On the 2D and 3D high-resolution seismic and GPR images of the Alpine Fault zone, we see the principal fault dipping steeply through Quaternary sediments and offsetting the basement. A distinct ~25 m vertical offset of basement provides a maximum ~1.4 mm/yr dip-slip displacement rate. The more important strike-slip component of displacement has yet to be estimated at this location. Our high-resolution seismic and GPR sections across parts of the northwest Canterbury Plains display a complex pattern of faults and folds beneath a variably thick veneer of flat-lying sediments. Structural restorations of the seismic images suggest 10 - 23% compressive strain, which would correspond to an average strain rate of 20 - 50 x 10-9/yr if the onset of compression coincided with the accelerated uplift of the Southern Alps approximately 5 Ma. Finally, multiple 2D high-resolution seismic images of the Ostler Fault zone reveal a 45 - 55 west-dipping principal fault and two subsidiary 25 - 30 west-dipping faults, one in the hanging wall and one in the footwall of the principal fault. Again, we are able to structurally restore models based on the seismic images. These restorations are compatible with 440 - 800 m of vertical offset and 870 - 1080 m of horizontal shortening across the Ostler Fault zone, which translate to a relatively constant deformation rate of 0.3 - 1.1 mm/yr since the Late Pliocene - Pleistocene.

Keywords: 2D and 3D high-resolution seismic reflection and GPR images, Australian - Pacific plate boundary, Alpine Fault zone, faults underlying the northwest Canterbury Plains, Ostler Fault zone. Introduction New Zealand's South Island is a dynamic landmass distinguished by two active subduction zones and an intervening transform plate boundary. The west-directed Hikurangi subduction zone in the north is connected to the east-directed Puysegur subduction zone in the south via the Alpine Fault zone and a broad expanse of oblique compression (Figure 1). Paleoseismological studies suggest the Alpine Fault is accommodating two-thirds to three-quarters of the 35 - 38 mm/y of relative motion between the Australian and Pacific plates in the central and southern parts of the South Island, with the remainder being distributed over a 150 - 200-kmwide region that encompasses the Southern Alps and adjacent regions to the east (Figure 1; Norris et al., 1990; Pettinga et al., 2001; Sutherland et al., 2006; Cox and Sutherland, 2007; Norris and Cooper, 2007). The Alpine Fault became a significant plate boundary 22 - 23 Ma (Cooper et al., 1987; King, 2000; Cox and Sutherland, 2007), about the same time the Hikurangi subduction zone started to consume the Pacific Plate (Stern et al., 2006). Since then, the zone of deformation has gradually broadened, with accelerated levels of faulting, folding, tilting, and uplift in the Southern Alps commencing ~5.0 Ma (Sutherland, 1995; Coates, 2002). The zone of deformation extended further east into the region of the Ostler Fault zone prior to the Late Pliocene and into the northwest Canterbury Plains either prior to or during the Quaternary (Forsyth et al., 2008). There are areas of active and Quaternary deformation close to the South Island's two largest cities, Christchurch and Dunedin (Figure 1) with populations of ~370,000 and ~124,000, respectively. The identification and characterization of active faults are critical to studies of regional seismic hazard (McCalpin, 1996; Yeats et al., 1996). Probabilistic seismic hazard analyses require seismic source models, which are usually constructed by assigning earthquakes to active faults based on seismological, other geophysical, and geological information. Once all seismic sources are defined, their characteristic earthquake magnitudes and recurrence frequencies are determined (Thenhaus and Campbell, 2003). Historical and instrumental records of seismicity do not completely characterize the earthquake cycle of many active fault zones, because the records are generally much shorter than the repeat times of the largest earthquakes. This is a particular problem in New Zealand, where relatively complete records of historical and instrumental seismicity exist for only the past ~200 and ~50 years, respectively, whereas large active faults on the islands are estimated to have recurrence intervals of many hundreds of years (Stirling et al, 1998). Consequently, paleoseismic studies are important for characterizing the numerous active faults that extend along the length of the country (e.g., Berryman, 1980; Van Dissen and Berryman, 1996; Benson et al., 2001; Villamor and Berryman, 2001; Sutherland et al., 2007). Conventional paleoseismological techniques based on surface outcrops, geomorphology, trenches, and boreholes (McCalpin, 1996; Yeats et al., 1996) usually supply details on major faults that have disrupted the upper few meters of the ground. To provide structural and other information at greater depths, we have acquired high-resolution seismic reflection and groundpenetrating radar (GPR) data across two active fault zones and an active fault network on the South Island. Our first target (1 in Figure 1) is the Alpine Fault zone at a rare accessible site where a ~2-m-high linear fault scarp cuts across a sequence of abandoned river terraces. Apart from a nearby campground, several houses, and a small village, this location in the north of the South Island is relatively isolated. A network of faults hidden beneath variably thick Quaternary 2

sediments that characterize the remarkably flat northwest Canterbury Plains is our second target (2 in Figure 1). A major buried or blind fault beneath the southernmost part of this target area is less than 40 km from Christchurch. The third target (3 in Figure 1) is a region of the Ostler Fault zone where thick sediments of the Mackenzie Basin are offset. This location in the south-central part of the Island is quite close to eight hydro-electric power dams that supply a large proportion of the South Island's electricity. After briefly describing the geology at the study sites, we show examples of the highresolution seismic reflection images at all three locations.

Figure 1. Plate tectonic setting of New Zealand's South Island showing the west-directed Hikurangi subduction zone in the north, the east-directed Puysegur subduction zone in the south, and the Alpine Fault. Also shown are our study sites: 1 - northern section of the Alpine Fault zone, 2 - northwest Canterbury Plains, 3 - Ostler Fault zone.

Geology in the vicinity of the three study sites Alpine Fault zone The Alpine Fault is the principal element of the onshore boundary between the Pacific and Australian plates (Figure 1). Offset of basement terrains by ~470 km along the fault since the late Oligocene to early Miocene is evidence for the importance of this structure in accommodating relative plate motion. At our Alpine Fault study site, a sequence of faulted abandoned river terraces is derived from glaciofluvial valley sediments dating from periods of aggradation that followed phases of advance and retreat of a late Pleistocene glacier. Basement southeast of the fault is schist derived from Triassic-age sediments, whereas that to the northwest is Paleozoic-age marble. Paleoseismic trenches on the youngest river terrace intercept a narrow zone of variably dipping (40o - 90o) faults within the upper 1 - 2 m of gravels (Yetton, 2002). Such gravels are either exposed at the surface or covered by a thin layer of undisturbed topsoil. GPR surveys at the site reveal a narrow fault zone dipping steeply to the southeast at ~80o to a depth of ~15 m (McClymont et al., 2008a, 2010). Notable changes in GPR reflection geometry are observed across the main strand of the fault. Additional changes in reflection characteristic ~30 m on either side of the main strand are interpreted as secondary left-stepping en echelon fault strands and associated deformation. This pattern of left-stepping traces is a consistent feature of the surface geomorphology within 1 km to the the SW of the site. The relationship between the main fault and secondary faults below the sediment cover is unknown. Northwest Canterbury Plains Lithologies and structures beneath the surficial sediments of the 8000 km2 Canterbury Plains are only poorly known. The landscape is flat and even in most areas, with little surface evidence of underlying geological complexity. Limited lithological information is provided by four moderately deep hydrocarbon exploration boreholes, a large number of shallow water-well holes, and outcrops in the hills and along the river banks (McLennan, 1981; Cowan, 1992; Evans, 2000; Estrada, 2003; May, 2004; Finnemore, 2004; Forsyth et al., 2008). Useful structural details are supplied by these same outcrops, sparse hydrocarbon exploration seismic lines (recorded with the intention of imaging much deeper features; Jongens et al., 1999), and a number of short seismic lines recorded by the University of Canterbury geophysics group (Estrada, 2003; Finnemore, 2004). Collectively, these sources of information suggest that the subsurface beneath the Canterbury Plains comprises Permian - Triassic basement rocks overlain successively by Late Cretaceous - Tertiary interbedded sedimentary and volcanic layers and Quaternary sediments. However, the details of most buried structures are either unknown or based on speculative extrapolations. Ostler Fault zone The Ostler Fault zone is one of a number of distinct structures east of the Alpine Fault that accounts for a significant component of strain between the two plates. It accommodates 1.1 - 1.7 mm/yr of east - west compression (Amos et al. 2007) and 0.8 - 1.0 mm/yr of vertical deformation that is mostly taken up by buckling in the fault hangingwall (Blick et al., 1989). In the area of the Mackenzie Basin transected by the Ostler Fault zone, a substantial late Pliocene - Pleistocene fluviolacustrine sequence is overlain by a series of Holocene glacial and fluvioglacial gravels and associated terraces (Ghisetti et al. 2007). The north - south striking Ostler Fault zone, which can be traced for a distance exceeding 50 km, comprises a highly segmented series of predominantly west-dipping surface-rupturing pure reverse faults. Fault segments are nearly orthogonal to inferred maximum horizontal-stress directions. They occur 4

over an area up to 3 km wide, with complex arrays of multiple fault segments in the central parts of the fault zone, where basin fill is thicker. The surface dip of the principal Ostler Fault plane appears to vary along its length. Davis et al. (2005) measured an average westerly dip of 50 9. A GPR survey in the northern part of the Clearburn area revealed a complex pattern of linked faulting, with multiple strands dipping to the west at 20 - 30 (McClymont et al. 2008b). Another 3D GPR survey in the northern section of the Ostler Fault zone found the principal fault plane to be westerly dipping at 51 10 (Amos et al., 2007; Wallace et al., 2010). Seismic reflection data acquisition and processing For each study site we designed a target-specific data acquisition strategy and a data-driven processing scheme. General information on the data acquisition strategies, all of which involved a 240-channel recording system and single 30 Hz geophones, are provided in the sections below, but we refer the readers to the relevant papers for details on the individual processing schemes (Kaiser et al., 2009, 2010; Dorn et al., 2010a, 2010b; Campbell et al., 2010a, 2010b, 2010c). After assigning coordinates and editing traces, we tested a full range of processing procedures on all data sets: (i) mutes of various types, including airwave attenuation, (ii) amplitude enhancement and trace balance, (iii) spectral balance, (iv) surface-consistent predictive deconvolution, (v) elevation statics, (vi) refraction statics with complementary P-wave velocity tomograms derived from first arrivals, (vii) surface-consistent residual statics, (viii) various types of velocity analysis, (ix) normal-moveout corrections, (x) a variety of dip-moveout (DM0) analyses and corrections, (xi) frequency filtering, (xii) frequency-wavenumber filtering, (xiii) frequency-space filtering, (xiv) various types of stacking, (xv) time migration followed by timeto-depth conversion, and (xvi) depth migration. The processes were not necessarily applied in the order listed and not all processes were successful in improving image quality. For example, application of refraction static corrections was essential for imaging the Alpine Fault, but resulted in markedly inferior images of the northwest Canterbury Plains and Ostler Fault regions. In contrast, DMO processing improved the images for the latter two regions, but yielded variable quality images across the Alpine Fault. Several of the processes were applied both pre- and poststack and some processes were applied in an iterative fashion (e.g., surface-consistent residual statics and velocity analyses). Results of seismic surveying the Alpine Fault zone Across the Alpine Fault zone, we recorded an ultra-high-resolution 2D seismic reflection data set along a 360-m-long line and a pseudo-3D seismic reflection data set that spanned a 500 x 200 m area (Kaiser et al., 2009, 2010). Energy generated by 6 blows of a hammer at 1 m intervals recorded on receivers located every 0.5 m provided 60-fold ultra-high-resolution data. For the pseudo-3D survey, energy from small 30 g explosive charges detonated every 8 m was recorded simultaneously by receivers located every 4 m along two parallel lines separated by 8 m; one of the two receiver lines was also the source line. This strategy provided >20-fold data on a uniform 2 x 4 m CMP grid. Together, these two data sets yielded very high-resolution details along a single profile and reliable information over a large area in a cost effective manner. Figures 2 and 3 show a depth-converted time-migrated 2D ultra-high-resolution image and a depth-migrated 3D high-resolution image of structures in the vicinity of the principal Alpine Fault (AF in the figures) strand. Seismic units A1, A2, and A3 of strong semicontinuous reflections on either side of the fault from near the surface to as deep as ~60 m are interpreted to be late Pleistocene gravels; trenches and pits at the site reveal alluvial gravel in the uppermost 5

2 m. Undifferentiated gravel was intersected from the near surface to 26 m depth in the borehole southeast of the surface fault trace (Figure 2b; Garrick and Hatherton, 1974); note the good correlation between the thickness of gravels observed in the seismic images and the borehole information. Along the southeastern half of the two data sets, reflections from seismic unit A3 appear to grade into gently dipping continuous reflections B0 - B2 at greater depth. We interpret seismic units B0 - B2 as originating from layered glaciolacustrine sediments.

Figure 2. (a) Depth-converted migrated section derived from the ultra-high-resolution seismic data set recorded across the Alpine Fault (AF). (b) Interpretation of depth-converted migrated section. The various units are described in the text. Information from a nearby ~80 m deep borehole is projected on to the line at about 143 m distance.

The sediment-basement interface produces strong reflections C1 and C2 that are offset across the Alpine Fault AF and at other locations along the profile. From mapped surface outcrops, we infer that marble basement C1 to the northwest is juxtaposed against schist basement C2 to the southeast. In Figures 2 and 3, much of the reverse slip along the Alpine Fault at Calf Paddock appears to be confined to a single major fault strand. More minor faults in Figure 2 are identified by dashed lines, with the most significant labeled I and II. Minor fault II projects to one of the subsidiary faults(SF in Figure 2) mapped in GPR data (McClymont et al., 2010). 6

Based on reflection truncations, the principal strand of the Alpine Fault has an apparent southeasterly dip of 7580 from its surface scarp through the Quaternary sediments to the offset basement at ~60 m depth. The dip of the Alpine Fault within the shallow basement is not so well constrained. As a consequence, we show a range of plausible depth trajectories for the Alpine Fault in Figure 2, with possible dip estimates in the basement varying between 50 and 80. Because reflections B0 - B2 from the former flat-lying glaciolacustrine units are only gently tilted, our preferred interpretation is that the Alpine Fault continues to dip relatively steeply within the shallow basement (Figure 3). The average ~25 m apparent vertical offset of basement across the principal Alpine Fault strand together with the estimated maximum age of the eroded basement surface and the measured fault dip within the Quaternary sediments yield a maximum dip-slip strain rate of ~1.4 mm/yr (estimated range of 0.9 2.2 mm/yr). The much larger strikeslip displacement rate has yet to be determined for this region of the Alpine Fault zone.

Figure 3. (a) Depth-migrated volume of the pseudo-3D data set recorded across the Alpine Fault. (b) As for (a), but highlighting the Alpine Fault strand AF and geological units A1-C2 discussed in the text. (c) 3D representation of the horizons picked in b). Dashed lines indicate where horizons onlap basement.

Results of seismic surveying across the northwest Canterbury Plains We recorded high-resolution seismic reflection data along 4 lines within the northwest Canterbury Plains (S1- S4 in Figure 4a - 4e; Dorn et al., 2010a, 2010b; Campbell et al., 2010b) using ~60 g explosive charges in 1-m-deep shotholes every 10 m. By recording the resultant energy on receivers every 5 m, we acquired 60-fold data at an interval of 2.5 m. Extensive rock exposures to the north and west of our study area and isolated outcrops to the east (Figure 4e) provide important constraints on the interpretation of the various seismic reflection patterns. The most important geological units for this study are: (i) unsaturated, largely unconsolidated young Quaternary gravel and sand, (ii) saturated, consolidated older Quaternary gravel and sand, (iii) interbedded Late Cretaceous - Tertiary sediments and volcanics, and (iv) Permian - Triassic Torlesse basement rocks. Additional critical information is supplied by refraction velocity tomograms derived from inversions of the first-arrival traveltimes recorded on the raw seismic data. Velocities in these tomograms vary from less than 1200 m/s in the shallow subsurface through to 3000 m/s and greater than 3500 m/s. By combining this information, we discover that the uppermost quasi-continuous subhorizontal reflections with velocities less than 1200 m/s originate from young Quaternary gravel and sand. The next deeper unit with reflections distinguished by distinct offsets and folding and velocities up to 3000 m/s is likely to be the older Quaternary gravel and sand, whereas the underlying unit with discontinuous to laterally continuous reflectivity and velocities up to 3000 m/s is very probably interbedded Late Cretaceous - Tertiary sediments and volcanics. Finally, reflection-free regions with velocities greater than 3500 m/s represent tightly folded and steeply dipping basement rocks. The interpreted seismic reflection images in Figure 4a - 4d are based on an integrated analysis of all available information. Abrupt truncations of prominent layered reflections from the Late Cretaceous - Tertiary and older Quaternary sequences and juxtaposition of reflections from these units with contrasting dips delineate a number of important faults. Moderate southeasterly dips and distinct curvature of the layered reflections define regions of tilting and folding. The vast majority of interpreted faults are distinguished by an anticlinal fold in the hanging wall and a synclinal fold in the footwall, both features being indicative of normal drag during reverse faulting (Grasemann et al., 2005). Of the 8 identified primary faults, 5 are plausible or probable extensions of exposed faults in the neighboring hills (i.e., FA1, FA2, FA2a, FA3a, and FA6). In particular, FA6 at the southern end of seismic section S2 (Figure 4c) may be a major basin bounding fault. A major fault-propagated fold (FO) recognized on the same seismic section can be confidently traced to an exposed anticline. The seismic sections also provide images of numerous secondary faults (thin dashed lines). Restorations to the top of the basement on the four longest seismic lines provide evidence for 10 - 23% compressive strain (Campbell et al., 2010b). This corresponds to an average strain rate of (i) 4.5 - 11.0 x 10-9 /yr if deformation commenced when movement along the Alpine Fault was initiated 23 Ma or (ii) 20 - 50 x 10-9/yr if compression began when uplift of the Southern Alps accelerated ~5 Ma. The strain appears to be spatially concentrated in the northern area of our study site. The geometry of the Quaternary units suggests that deformation is ongoing in this region. A small discrete offset of reflections from very recent sediments (not shown here, Campbell et al., 2010b), weak curvature of reflections from the youngest Quaternary sediments, and minor undulations in the surface topography (May, 2004, Forsyth et al., 2008) are evidence for blind or buried active faults beneath some of the seismic lines.

Figure 4. High-resolution seismic reflection data recorded across the northwest Canterbury Plains. Sketch of interpreted geology superimposed on timemigrated seismic sections for (a) S4, (b) S1, (c) S2 and (d) S3. Solid and dashed lines indicate the level of confidence in the interpreted features. The very approximate depth scale on the right is based on a single velocity of 2300 m/s. (e) Location of the S1 - S4 seismic lines in the northwest Canterbury Plains. Regions of rock outcrop are shaded.

Results of seismic surveying across the Ostler Fault zone Our high-resolution seismic reflection data in the region of the Ostler Fault zone spanned a transfer zone from one principal fault strand to another. Six pairs of subparallel 1.2-km-long seismic reflection lines were recorded orthogonal to the fault traces and a single 1.6-km-long crossline was collected in the hanging wall parallel to fault strike (see top right diagram of Figure 5). Since time and financial constraints prevented the acquisition of a sufficiently high-resolution 3D data set across the 1.2 x 1.6 km survey area, the seismic survey was designed to supply highresolution cross-sections through the predominantly 2D structures, while providing constraints on the along-strike character of these structures at length scales of both ~50 m (between lines in each pair) and 200 - 300 m (between adjacent pairs of lines). The pairs of lines gave us confidence that the vast majority of reflected energy seen on the seismic images originated from within the vertical planes defined by the respective recording lines. The acquisition parameters were very similar to those described for the northwest Canterbury Plains data, except the source, receiver and fold were 6, 3, and 1.5 m, respectively, and the shotholes were generally shallower at ~0.6 m. The upper two diagrams in the left of Figure 5 show depth-converted migrated sections from inline 3 and the crossline (Campbell et al., 2010a, 2010c). Nearby rock outcrops allow us to identify the most prominent reflections. The shallowest subhorizontal events P - P' and R - R' originate from fluvioglacial gravel beds, whereas the high-amplitude dipping reflections Q - Q' and deeper subhorizontal reflections U - U' originate from layered Late Pliocene - Pleistocene fluviolacustrine units (Ghisetti et al., 2007). We interpret T - T' as a rare package of reflections from a fault. Correlation of reflections, reflection packages, and reflection truncations on all 13 seismic sections suitably constrained by the outcrop data leads to the interpretation presented in the lower two diagrams of Figure 5. Major unknowns in our interpretation are the types of rock underlying the deepest imaged reflections U - U'. Nevertheless, it is possible to interpret the seismic images in terms of a principal 45 - 55 west-dipping Ostler Fault plane running the entire length of the study site and two subsidiary 25 - 30 west-dipping faults, one in the hanging wall and one in the footwall of the principal fault. The principal fault and its shallow splays offset the Late Pliocene to Quaternary units. Deformation style varies along the length of the site, with folding playing an increasingly important role towards the north. Structural restorations of inline sections (Campbell et al., 2010a) demonstrate that 440 - 800 m of vertical strain and 870 - 1080 m of shortening has occurred since deposition of the earliest Late Pliocene - Pleistocene fluviolacustrine unit (L1 in Figure 5). Furthermore, they suggest that there has been 70 - 180 m of shortening since the start of Quaternary fluvioglacial deposition. These estimates together with poorly constrained timing information translate to a relatively constant deformation rate of ~0.3 - 1.1 mm/yr from the Late Pliocene - Pleistocene to the present day. Conclusions Ultra-high-resolution and high-resolution seismic reflection sections and 3D GPR data (not presented here) have revealed important details on the structure of two major active fault zones and a network of potentially active faults on New Zealand's South Island. We have imaged the transpressive plate-boundary Alpine Fault in 2D and 3D and reverse faults of the Ostler Fault zone and the region underlying the northwest Canterbury Plains.

10

Figure 5 High-resolution seismic reflection data recorded across the Ostler Fault. Top right diagram shows the field layout. Six pairs of parallel lines were recorded perpendicular to the fault strike (indicated by the lineations in the topography) and one was recorded perpendicular (i.e., the crossline). Depth-converted migrated section 3 and the crossline (both marked by heavier lines in the top right diagram) are shown in the top left and center diagrams. Interpreted versions of these sections are shown below.

11

Acknowledgments We thank all members of the field crews who worked so hard on our surveys. This project was funded by the Swiss National Science Foundation and ETH Zurich. References Amos, C.B., Burbank, D.W., Nobes, D.C., and Read, S.A.L., 2007, Geomorphic constraints on listric thrust faulting: Implications for active deformation in the Mackenzie Basin, South Island, New Zealand: Journal of Geophysical Research, 112, B03S11. Benson, A.M., Little, T.A., Van Dissen, R.J., Hill, N., and Townsend, D.B., 2001, Late Quaternary paleoseismic history and surface rupture characteristics of the eastern Awatere strike-slip fault, New Zealand: Geological Society of America Bulletin, 113, 1079-1091. Berryman, K.R., 1980, Late Quaternary movement of the White Creek Fault, South Island, New Zealand: New Zealand Journal of Geology and Geophysics, 23, 93-101. Blick, G.H., Read, S.A.L., and Hall, P.T., 1989, Deformation monitoring of the Ostler fault zone, South Island, New Zealand: Tectonophysics, 167, 329-339. Campbell, F.M., Ghisetti, F., Kaiser, A.E., Green, A.G., Horstmeyer, H., and Gorman, A.R., 2010a, Structure and evolution of the seismically active Ostler Fault Zone (New Zealand) based on interpretations of multiple high-resolution seismic reflection profiles: Journal of Structural Geology, submitted. Campbell, F., Jongens, R., Green, A.G., Dorn, C., Carpentier, S., Kaiser, A.E., Horstmeyer, H., Campbell, J., Finnemore, M., and Pettinga, J., 2010b, Mapping and restoring Late Cretaceous - Present potentially seismogenic structures beneath the northwest Canterbury Plains (New Zealand) based on high-resolution seismic reflection data and geological information: Tectonics, submitted. Campbell, F.M., Kaiser, A., Horstmeyer, H., Green, A., Ghisetti, F., Gorman, A.R., Finnemore, M., and Nobes, D.C., 2010c, Processing and preliminary interpretation of noisy highresolution seismic reection/refraction data across the active Ostler Fault zone, South Island, New Zealand: Journal of Applied Geophysics, in press. Coates, G., 2002, The Rise and Fall of the Southern Alps: Canterbury University Press, Christchurch. Cooper, A. F., Barreiro, B.A., Kimbrough, D.L., and Martinson, J. M., 1987, Lamprophyre dike intrusion and the age of the Alpine Fault, New Zealand: Geology, 15, 941-944. Cowan, H., 1992, Structure, seismicity and tectonics of the Porter's Pass-Amberley fault zone, North Canterbury, New Zealand: PhD Thesis, University of Canterbury, Christchurch. Cox, S.C., and Sutherland, R., 2007, Regional geological framework of South Island, New Zealand, and its significance for understanding the active plate boundary: American Geophysical Union Geophysical Monograph, 175, 19-46. Davis, K., Burbank, D.W., Fisher, D., Wallace, S., and Nobes, D., 2005, Thrust-fault growth and segment linkage in the active Ostler fault zone, New Zealand: Journal of Structural Geology, 27, 1528-1546.

12

Dorn, C., Carpentier, S., Kaiser, A.E., Green, A.G., Horstmeyer, H., Campbell, F., Campbell, J., Jongens, R., Finnemore, M., and Nobes, D.C., 2010a, First seismic imaging results of tectonically complex structures at shallow depths beneath the northwest Canterbury Plains, New Zealand; Journal of Applied Geophysics, in press. Dorn, C., Green, A.G., Jongens, R., Carpentier, S., Kaiser, A.E., Campbell, F., Horstmeyer, H., Campbell, J., Finnemore, M., and Pettinga, J., 2010b, High-resolution seismic images of potentially seismogenic structures beneath the northwest Canterbury Plains, New Zealand: Journal of Geophysical Research, submitted. Estrada, B., 2003, Seismic hazard associated with the Springbank Fault, North Canterbury Plains. MSc Thesis, University of Canterbury, Christchurch. Evans, S.T., 2000, Paleoseismic analysis of the Springfield Fault, Central Canterbury: BSc Thesis, University of Canterbury, Christchurch. Finnemore, M., 2004, The application of seismic reflection surveying to the characterisation of aquifer geometry and related active tectonic deformation, North Canterbury: PhD Thesis, University of Canterbury, Christchurch. Forsyth, P.J., Barrell, D.J.A., and Jongens, R. (compilers), 2008, Geology of the Christchurch area: Institute of Geological and Nuclear Sciences 1:250 000 Geological Map and Report 16: GNS Science, Lower Hutt. Garrick, R. A., and Hatherton, T., 1974, Seismic refraction profiles across the Alpine Fault: Report 87, Geophysics Division, Department of Science and Industry Research, Wellington. Ghisetti, F.C., Gorman, A.R., and Sibson, R.H., 2007, Surface breakthrough of a basement fault by repeated seismic slip episodes: The Ostler Fault, South Island, New Zealand; Tectonics, 26, TC6004. Grasemann, B., Martel, S., and Passchier, C., 2005, Reverse and normal drag along a fault: Journal of Structural Geology, 27, 9991010. Jongens, R., Pettinga, J.R., Campbell, J.K., 1999. Stratigraphic and structural overview of the onshore Canterbury basin: North Canterbury to the Rangitata River: Unpublished Open-File Petroleum Report PR4067, Ministry of Economic Development, Wellington. Kaiser, A.E., Green, A.G., Campbell, F.M., Horstmeyer, H., Manukyan, E., Langridge, R.M., McClymont, A.F., Mancktelow, N., Finnemore, M., and Nobes, D.C., 2009, Ultra-highresolution seismic reflection imaging of the Alpine Fault, New Zealand: Journal of Geophysical Research, 114, B11306. Kaiser, A.E., Horstmeyer, H., Green, A.G., Campbell, F.M., Langridge, R.M., and McClymont, A.F., 2010, Detailed images of the shallow Alpine Fault Zone (New Zealand) determined from pseudo-3D seismic reflection data: Geophysics, submitted. King, P., 2000, Plate reconstructions for New Zealand: New Zealand Journal of Geology and Geophysics, 43, 611-638. May, B.D., 2004, Comparative geomorphology of two active tectonic structures near Oxford, North Canterbury: MSc Thesis, University of Canterbury, Christchurch. McCalpin, J.P., 1996, Paleoseismology: Academic Press, San Diego.

13

McClymont, A.F., Green, A.G., Streich, R., Horstmeyer, H., Tronicke, J., Nobes, D.C., Pettinga, J., Campbell, J., and Langridge, R., 2008a, Visualization of active faults using geometric attributes of 3D GPR data: An example from the Alpine Fault Zone, New Zealand: Geophysics, 73, B11B23. McClymont, A., Green, A.G., Villamor, P., Horstmeyer, H., Grass C., and Nobes, D., 2008b, Characterization of active faults using 3D GPR data: Journal of Geophysical Research, 113, B10315. McClymont, A.F., Green, A.G., Kaiser, A., Horstmeyer, H., and Langridge, R., 2010, Shallow fault segmentation of the Alpine fault zone, New Zealand revealed from 2- and 3-D GPR surveying: Journal of Applied Geophysics, in press. McLennan, J.M., 1981, The Cretaceous-Tertiary rocks of Avoca, Oxford and Burnt Hill, Central Canterbury: MSc Thesis, University of Canterbury, Christchurch. Norris, R.J., Koons, P.O., and Cooper, A.F., 1990, The obliquely-convergent plate boundary in the South Island of New Zealand: implications for ancient collision zones: Journal of Structural Geology, 12, 715-725. Norris, R.J., and Cooper, A.F., 2007. The Alpine Fault, New Zealand; surface geology and field relationships: American Geophysical Union Geophysical Monograph, 175, 157175. Pettinga, J.R., Yetton, M.D., Van Dissen, R.J., and Downes, G., 2001, Earthquake source identification and characterisation for the Canterbury Region, South Island, New Zealand: Bulletin of the New Zealand Society for Earthquake Engineering, 34 (4), 282-315. Stern, T. A., Stratford, W.R., and Salmon, M. L., 2006, Subduction evolution and mantle dynamics at a continental margin: central North Island, New Zealand: Reviews of Geophysics, 44, RG4002. Stirling, M.W., Wesnousky, S.G., and Berryman, K.R., 1998, Probabilistic seismic hazard analysis of New Zealand: New Zealand Journal of Geology and Geophysics, 41, 355-375. Sutherland, R., 1995, The Australia-Pacific boundary and Cenozoic plate motions in the SW Pacific: Some constraints from Geosat data: Tectonics, 14, 819-831. Sutherland, R., Berryman, K., and Norris, R., 2006, Quaternary slip rate and geomorphology of the Alpine fault: Implications for kinematics and seismic hazard in southwest New Zealand: Geological Society of America Bulletin, 118, 464-474. Sutherland, R., Eberhart-Phillips, D., Harris, R.A., Stern, T., Beavan J., Ellis, S., Henrys, S., Cox, S.C., Norris, R.J., Berryman, K.R., Townend, J., Bannister, S., Pettinga, J., Leitner, B., Wallace, L., Little, T.A., Cooper, A.F., Yetton, M., and Stirling, M., 2007, Do great earthquakes occur on the Alpine fault in central South Island, New Zealand? American Geophysical Union Geophysical Monograph, 175, 235 251. Thenhaus, P. C., and Campbell, K.W., 2003, Seismic hazard analysis: In Earthquake Engineering Handbook, edited by W. F. Chen and C. Scawthorn, 1 50: CRC Press, Boca Raton. Van Dissen, R.J., and Berryman, K.R., 1996, Surface rupture earthquakes over the last ~1000 years in the Wellington region, New Zealand, and implications for ground shaking hazard: Journal of Geophysical Research, 101, 5999-6019.

14

Villamor, P., and Berryman, K.R., 2001, A late Quaternary extension rate in the Taupo Volcanic Zone, New Zealand, derived from fault slip data: New Zealand Journal of Geology and Geophysics, 44, 243-269. Wallace, S.C., Nobes, D.C., Davis, K.J., Burbank, D.W., and white, A., 2010, Three-dimensional GPR imaging of the Benmore anticline and step-over of the Ostler Fault, South Island, New Zealand, Geophysical Journal International, 180, 465474. Yeats, R.S., Sieh, K.E., and Allen, C.R., 1996, Geology of earthquakes: Oxford University Press, Oxford. Yetton, M.D., 2002, Paleoseismic investigation of the North and West Wairau sections of the Alpine Fault, South Island, New Zealand: Earthquake Commission Research Foundation, Report 99/353, Christchurch.

15